hey Juk,

will you latest version (or any versions) work for Stars 18 mans?

No sorry, it'll only work for SNGs with 10 or less players atm.

Juk

back to work for yourself so :P

not the space but the pot size was the problem. That one is already solved, but here is another one:

#Game No : 2297368-17
***** Hand History for Game 2297368-17 *****
NL Texas Hold'em $11 USD Buy-in Trny:2297368 Blinds(75/150) - Thursday, 14:47:57
Seat 9 is the button
Seat 2: stykk4e890 ( 35 )
Seat 4: RaisingSun ( 2685 )
Seat 9: Threepwood ( 4780 )
** Dealing down cards **
Dealt to Threepwood [ Ac Tc ]
Threepwood all-in [4780]
RaisingSun all-in [2535]
** Dealing Flop ** [ 2s, 3d, 7h ]
** Dealing Turn ** [ Qc ]
** Dealing River ** [ Ks ]
stykk4e890 wins 105 chips from the main pot
Threepwood wins 7395 chips from the side pot #1
stykk4e890 shows [ 4d, Kd ]
RaisingSun shows [ Ad, 6c ]
Threepwood shows [ Ac, Tc ]

the parser doesnt find any problem and no error.log was created but something is wrong with the result: adjusted won EVLuck ect is 0.

if i correct it to "Threepwood wins 7500 chips" and delete the line "stykk4e890 wins 105 chips from the main pot" the result seems to be ok.

it looks like the problem is the side pot handling. do u have any idea how to fix this?

Is the final hand of this tournament also parsed (correctly) after this one? That would explain why changing to "Threepwood wins 7500 chips" works (just running this hand on it's own or with a broken final hand of the SNG will cause this to get ignored).

Juk

Seems like a great program.

The link doesn't work for some reason.

This should work: http://www.jukofyork.com/Juks_BackTe...uck_v1_03e.rar

Not sure why - I've re-entered the URL in the post so hopefully that fixes it.

Juk

Working now!

Thanks, juk and fievel



I'll check it out later today.

Okay, I've been using it for a bit. Had to tweak some stuff for FTP, but it works pretty well.

Question..... how do I make a graph out of the data? I downloaded the php5 rar, but I'm not sure what to do with the contents.

I have the two spreadsheets, too.

Thank you!

i'm tryng to use the program, but i didn't understood the output
what does this mean? why actual won is 0?

NumSNGs : 145 (NumDataPoints=1248)
> 2-way=1185, 3-way=70, multiway=4
> pre-flop=930, flop=211, turn=118
Actual Won : 0$ (Actual ROI=-1.#IND%)
Adjusted Won : 0$ (Adjusted ROI=-1.#IND%)
$EV Luck : 0$ (0$ per all-in)
cEV Luck : -43241t (-34.6489t per all-in)

Hello all,

I just downloaded v1 03e and tryed to import some hands. I get a couple of different errors:

(all files are 1 table/ 9seats/ regular/1st 2nd 3rd/5.50$/on stars) 94 files...


1. NewAction() FOLD failed... Skipping to next hand.] Continuing...

2. NewAction() RAISE failed... Skipping to next hand.] Continuing...

3. NewAction() CALL failed... Skipping to next hand.] Continuing...

4. Player acted out-of-turn.] Continuing...

5. FATAL: [Too many chips in play to use short data-type in HoldemGameState.] Stopping...

6. BB must be >0 in NewHand().] Continuing...

7. Could not find the hour of the day... Skipping to next hand.] Continuing...

8. [|Can't find the '/' in blind... Skipping to next hand.] Continuing...

any ideas what I am doing wrong here?

Thanks for hints

Not even for cEV?

The cEV stuff will work so long as the total amount of chips in play at a table isn't greater than about 32k (no good reason for this limit - just a dodgy constant that came from using 16bit shorts to store the stack sizes...).

There are other tools that will work with cEV though (eg: SNGLuck).

Juk

Yeah I ran into that limitation on my test 5 minutes ago!!

Thanks for the link! Will try that!

would it be possible to expand on this to also calculate the luck impact of other players all-ins?

For example, QQ and a shorter stack with A6 get into it on the bubble, A6 gets there. Obviously thats a pretty unlucky result for us. Think its just the same ICM calculation, except looking at our $EV for each result instead of the other players.

Would be nice to see that luck factored in as well (seperate line, then seperate total?). Just something else to cross out in order to get to true ROI faster.

Other player's all-ins are already accounted for in the current code (it's not drawn as a separate line though).

Juk

Hey all, I read through the first 10 pages and saw that the early Beta's didn't support Pokerstars. I admit I did not want to sift through 80 pages for an answer that is easily told to me.

Does anyone know of a version that is compatible with Pokerstars? Or has this one been updated for it? Specifically for Double or Nothing sit n' go's. I actually am not looking for an equity for my all-ins based upon ICM statistics and relative chip stacks, players left, etc. according to the pay out structure.

All I care to know is what were my pre-flop all-in chances of winning the hand against the villain's hand that I faced? That's it. Plain and simple. Does this system break it down into hand categories as well? For example Pocket Pair vs lower PP, AK vs two unders, A-small vs two cards in the middle (ex: K-J), 2 to 1 favorite hands like QQ vs A-7, and finally hands where you have someone dominated or are dominated (ex: KQ vs K-10)?

These are the types of hands I'm extremely curious to know how I'm running in. The format of tournament I play is so repetitive. I am on a MASSIVE downswing, but am an overall winning player in the format. I am curious to upload my last weeks worth of HH into a Luck analyzer to compute some findings as listed above. If anyone knows of one that does or doesn't require Pokertracker I would be much oblidged if you'd fill me in on it. I don't use PT so if it does, I'd pay someone through a transfer on PS around $10 or $20 depending on how much work is involved with making the graphs, etc. for me. Thanks again, feel free to PM me! Thanks!

My compliments to the original creator.

Ahhh nice, wow I've used it for nine months and never realized

Is there any way that the $EV is being calculated incorrectly since the big update at FT? Since that update it is telling me that I am running nearly 300 buy-ins under equity but still winning at like 4% ROI. I think I'm good, but I'm not sure if anybody is THAT good. I haven't had a good EV run since the update, which was about 7,000 sngs for me.

Do you still have an old version (I dont) and a sample of old hands?

Can just run old and new versions on that sample then...

The only problem is that the old version doesn't work post-update. I did run the new version on pre-update hands in a sng stretch where I knew what my adjusted equity should be and it came out okay.

However, I find it kind of odd that the biggest -EV stretch that I've ever heard of at the super turbos from anybody was -200 buy-ins, and since the update, I have had a -220 buy-in EV graph at the 7.50s, where I still maintained a 5% roi, which was immediately followed by a -100 BI EV stretch at the $15s where I still have a 4% roi. Of that -100 buy-ins, half of it was in 1 day where I got absolutely demolished, but the other half was over the course of about 1,200 sngs where I felt like I was running really well but was showing 50 buy-ins under EV.

I understand that being on a bad run in the past has no bearing on what my future results will be, but generally luck changes course a bit more in 7k sngs than I have seen dictated by the updated analyzer.

I don't really know enough about programming to look at the way things are calculated, but I haven't had a significant EV upswing in over 7k games and there have been stretches where I felt like I was running pretty well and still showed a higher adjusted earnings than actual. I'd like to think that I am godlike despite running under EV, but after that many sngs I think it is more likely that there is an issue with the program.

The problem is most likely to be some bias with the FTP converter (ie: it's missing/misconverting certain hands).

Juk

Looks like converter have bug when using for big number of hands. File without tournament number or empty hands. I decide to fix it when I meet it again, but it never happens and no one report bugs.
Do you have errors during analysis? Try use smaller sample sizes. Try not mix hands in old and new format. I will see if I can find some bugs.
I run test on my last 1000 SNG and it works fine. So if you will find HH sample about 500 games that cause strange result, please send me to aruta[at]sibmail.com, I will try to solve problem.

Thanks Aruta. I am actually most concerned with my last 2-3k sngs, but I run a new graph for each individual session (usually 60-130 sngs), so it shouldn't be the actual # of hands or sngs that is the problem. Maybe the results are legit (I actually hope they are) and I'm just being paranoid. If they are legit, I've got a big upswing coming my way which I am excited about. I do get the occasional error, but usually it is just skipping 1 hand every once in a while.